diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11926 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "global_step": 1964054,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.0500635531186734e-05,
+      "loss": 7.9258,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4647181356188376e-05,
+      "loss": 6.7206,
+      "step": 2000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.705525479232612e-05,
+      "loss": 6.2705,
+      "step": 3000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.875604750285333e-05,
+      "loss": 5.9721,
+      "step": 4000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 5.7209,
+      "step": 5000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 5.5184,
+      "step": 6000
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-05,
+      "loss": 5.3619,
+      "step": 7000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 5.2279,
+      "step": 8000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 5.1094,
+      "step": 9000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 5.0125,
+      "step": 10000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 4.9219,
+      "step": 11000
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-05,
+      "loss": 4.8491,
+      "step": 12000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 4.7722,
+      "step": 13000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 4.7121,
+      "step": 14000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 4.6485,
+      "step": 15000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 4.582,
+      "step": 16000
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5e-05,
+      "loss": 4.5365,
+      "step": 17000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 4.4739,
+      "step": 18000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 4.4262,
+      "step": 19000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 4.3785,
+      "step": 20000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 4.3253,
+      "step": 21000
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 5e-05,
+      "loss": 4.2774,
+      "step": 22000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 4.2332,
+      "step": 23000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 4.1947,
+      "step": 24000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 4.1613,
+      "step": 25000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 4.1186,
+      "step": 26000
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 5e-05,
+      "loss": 4.0885,
+      "step": 27000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 4.0514,
+      "step": 28000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 4.019,
+      "step": 29000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 3.984,
+      "step": 30000
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 5e-05,
+      "loss": 3.9533,
+      "step": 31000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 3.9193,
+      "step": 32000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 3.9009,
+      "step": 33000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 3.8693,
+      "step": 34000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 3.8376,
+      "step": 35000
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 5e-05,
+      "loss": 3.8107,
+      "step": 36000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 3.7897,
+      "step": 37000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 3.7741,
+      "step": 38000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 3.7481,
+      "step": 39000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 3.726,
+      "step": 40000
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 5e-05,
+      "loss": 3.6989,
+      "step": 41000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 3.6819,
+      "step": 42000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 3.6614,
+      "step": 43000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 3.6406,
+      "step": 44000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 3.6174,
+      "step": 45000
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 5e-05,
+      "loss": 3.6004,
+      "step": 46000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 3.5884,
+      "step": 47000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 3.5653,
+      "step": 48000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 3.554,
+      "step": 49000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 3.5352,
+      "step": 50000
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 5e-05,
+      "loss": 3.5185,
+      "step": 51000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 3.5032,
+      "step": 52000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 3.4902,
+      "step": 53000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 3.4716,
+      "step": 54000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 3.4541,
+      "step": 55000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 5e-05,
+      "loss": 3.4419,
+      "step": 56000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 3.4297,
+      "step": 57000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 3.4116,
+      "step": 58000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 3.4023,
+      "step": 59000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 3.3892,
+      "step": 60000
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 5e-05,
+      "loss": 3.3774,
+      "step": 61000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 3.368,
+      "step": 62000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 3.3504,
+      "step": 63000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 3.3382,
+      "step": 64000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 3.3264,
+      "step": 65000
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 5e-05,
+      "loss": 3.3244,
+      "step": 66000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 3.3115,
+      "step": 67000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 3.3021,
+      "step": 68000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 3.287,
+      "step": 69000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 3.287,
+      "step": 70000
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 5e-05,
+      "loss": 3.2652,
+      "step": 71000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 3.2519,
+      "step": 72000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 3.2475,
+      "step": 73000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 3.239,
+      "step": 74000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 3.2336,
+      "step": 75000
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5e-05,
+      "loss": 3.2193,
+      "step": 76000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 3.2119,
+      "step": 77000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 3.2065,
+      "step": 78000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 3.1932,
+      "step": 79000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 3.1863,
+      "step": 80000
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5e-05,
+      "loss": 3.1714,
+      "step": 81000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 3.1721,
+      "step": 82000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 3.1595,
+      "step": 83000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 3.1564,
+      "step": 84000
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5e-05,
+      "loss": 3.1508,
+      "step": 85000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 3.131,
+      "step": 86000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 3.1199,
+      "step": 87000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 3.1231,
+      "step": 88000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 3.1172,
+      "step": 89000
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5e-05,
+      "loss": 3.1107,
+      "step": 90000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 3.1006,
+      "step": 91000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 3.0946,
+      "step": 92000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 3.0934,
+      "step": 93000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 3.0837,
+      "step": 94000
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 5e-05,
+      "loss": 3.074,
+      "step": 95000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 3.0712,
+      "step": 96000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 3.0704,
+      "step": 97000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 3.0613,
+      "step": 98000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 3.046,
+      "step": 99000
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 5e-05,
+      "loss": 3.0416,
+      "step": 100000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 3.0417,
+      "step": 101000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 3.0353,
+      "step": 102000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 3.0309,
+      "step": 103000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 3.0247,
+      "step": 104000
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 5e-05,
+      "loss": 3.0213,
+      "step": 105000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 3.0103,
+      "step": 106000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 2.9965,
+      "step": 107000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 2.9967,
+      "step": 108000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 2.9975,
+      "step": 109000
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 5e-05,
+      "loss": 2.9872,
+      "step": 110000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 2.982,
+      "step": 111000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 2.9821,
+      "step": 112000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 2.9704,
+      "step": 113000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 2.9698,
+      "step": 114000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 5e-05,
+      "loss": 2.9595,
+      "step": 115000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 2.9603,
+      "step": 116000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 2.9484,
+      "step": 117000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 2.9532,
+      "step": 118000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 2.9456,
+      "step": 119000
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 5e-05,
+      "loss": 2.9333,
+      "step": 120000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 2.9323,
+      "step": 121000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 2.9328,
+      "step": 122000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 2.9312,
+      "step": 123000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 2.9226,
+      "step": 124000
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 5e-05,
+      "loss": 2.9158,
+      "step": 125000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 2.9064,
+      "step": 126000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 2.9094,
+      "step": 127000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 2.8974,
+      "step": 128000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 2.8964,
+      "step": 129000
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 5e-05,
+      "loss": 2.8948,
+      "step": 130000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 2.8862,
+      "step": 131000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 2.8835,
+      "step": 132000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 2.8863,
+      "step": 133000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 2.8821,
+      "step": 134000
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 5e-05,
+      "loss": 2.8781,
+      "step": 135000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 2.8695,
+      "step": 136000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 2.8673,
+      "step": 137000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 2.8605,
+      "step": 138000
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 5e-05,
+      "loss": 2.865,
+      "step": 139000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 2.8576,
+      "step": 140000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 2.8584,
+      "step": 141000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 2.8458,
+      "step": 142000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 2.8427,
+      "step": 143000
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 5e-05,
+      "loss": 2.8444,
+      "step": 144000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 2.8391,
+      "step": 145000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 2.8311,
+      "step": 146000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 2.8212,
+      "step": 147000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 2.8259,
+      "step": 148000
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 5e-05,
+      "loss": 2.827,
+      "step": 149000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 2.824,
+      "step": 150000
+    },
+    {
+      "epoch": 0.31,
+      "eval_accuracy": 0.5210951936592287,
+      "eval_loss": 2.658203125,
+      "eval_runtime": 4374.358,
+      "eval_samples_per_second": 145.413,
+      "eval_steps_per_second": 1.136,
+      "step": 150000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 2.8211,
+      "step": 151000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 2.8055,
+      "step": 152000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 2.8112,
+      "step": 153000
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 5e-05,
+      "loss": 2.8067,
+      "step": 154000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 2.8039,
+      "step": 155000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 2.8047,
+      "step": 156000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 2.7934,
+      "step": 157000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 2.793,
+      "step": 158000
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 5e-05,
+      "loss": 2.7861,
+      "step": 159000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 2.7845,
+      "step": 160000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 2.7863,
+      "step": 161000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 2.7825,
+      "step": 162000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 2.7833,
+      "step": 163000
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 5e-05,
+      "loss": 2.7791,
+      "step": 164000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 2.7702,
+      "step": 165000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 2.77,
+      "step": 166000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 2.7637,
+      "step": 167000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 2.7632,
+      "step": 168000
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 5e-05,
+      "loss": 2.7622,
+      "step": 169000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 2.7558,
+      "step": 170000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 2.7611,
+      "step": 171000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 2.7467,
+      "step": 172000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 2.7524,
+      "step": 173000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 5e-05,
+      "loss": 2.743,
+      "step": 174000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 2.7393,
+      "step": 175000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 2.7372,
+      "step": 176000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 2.7393,
+      "step": 177000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 2.7414,
+      "step": 178000
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 5e-05,
+      "loss": 2.7389,
+      "step": 179000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 2.7251,
+      "step": 180000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 2.7306,
+      "step": 181000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 2.7324,
+      "step": 182000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 2.723,
+      "step": 183000
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 5e-05,
+      "loss": 2.7187,
+      "step": 184000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 2.7174,
+      "step": 185000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 2.7184,
+      "step": 186000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 2.7167,
+      "step": 187000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 2.7103,
+      "step": 188000
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 5e-05,
+      "loss": 2.7066,
+      "step": 189000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 2.7023,
+      "step": 190000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 2.7059,
+      "step": 191000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 2.6976,
+      "step": 192000
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 5e-05,
+      "loss": 2.6931,
+      "step": 193000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 2.6967,
+      "step": 194000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 2.6932,
+      "step": 195000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 2.6876,
+      "step": 196000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 2.6884,
+      "step": 197000
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 5e-05,
+      "loss": 2.6927,
+      "step": 198000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 2.6813,
+      "step": 199000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 2.6867,
+      "step": 200000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 2.6853,
+      "step": 201000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 2.6862,
+      "step": 202000
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 5e-05,
+      "loss": 2.6763,
+      "step": 203000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 2.6702,
+      "step": 204000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 2.6732,
+      "step": 205000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 2.6713,
+      "step": 206000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 2.6722,
+      "step": 207000
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 5e-05,
+      "loss": 2.6585,
+      "step": 208000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 2.6609,
+      "step": 209000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 2.6594,
+      "step": 210000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 2.652,
+      "step": 211000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 2.6545,
+      "step": 212000
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 5e-05,
+      "loss": 2.6552,
+      "step": 213000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 2.6618,
+      "step": 214000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 2.6521,
+      "step": 215000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 2.642,
+      "step": 216000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 2.6433,
+      "step": 217000
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 5e-05,
+      "loss": 2.6456,
+      "step": 218000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 2.6445,
+      "step": 219000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 2.6466,
+      "step": 220000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 2.6393,
+      "step": 221000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 2.6348,
+      "step": 222000
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 5e-05,
+      "loss": 2.6382,
+      "step": 223000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 2.6366,
+      "step": 224000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 2.6332,
+      "step": 225000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 2.631,
+      "step": 226000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 2.6286,
+      "step": 227000
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 5e-05,
+      "loss": 2.6269,
+      "step": 228000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 2.6244,
+      "step": 229000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 2.629,
+      "step": 230000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 2.6219,
+      "step": 231000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 2.6185,
+      "step": 232000
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 5e-05,
+      "loss": 2.6206,
+      "step": 233000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 2.618,
+      "step": 234000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 2.6179,
+      "step": 235000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 2.6116,
+      "step": 236000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 2.6078,
+      "step": 237000
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 5e-05,
+      "loss": 2.6169,
+      "step": 238000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 2.6091,
+      "step": 239000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 2.6065,
+      "step": 240000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 2.6078,
+      "step": 241000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 2.6051,
+      "step": 242000
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 5e-05,
+      "loss": 2.5986,
+      "step": 243000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 2.6005,
+      "step": 244000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 2.5998,
+      "step": 245000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 2.6003,
+      "step": 246000
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 5e-05,
+      "loss": 2.5952,
+      "step": 247000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 2.5907,
+      "step": 248000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 2.5934,
+      "step": 249000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 2.5948,
+      "step": 250000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 2.5879,
+      "step": 251000
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 5e-05,
+      "loss": 2.5868,
+      "step": 252000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 2.5885,
+      "step": 253000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 2.5854,
+      "step": 254000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 2.5853,
+      "step": 255000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 2.5803,
+      "step": 256000
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 5e-05,
+      "loss": 2.5759,
+      "step": 257000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 2.5729,
+      "step": 258000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 2.5737,
+      "step": 259000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 2.5786,
+      "step": 260000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 2.5704,
+      "step": 261000
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 5e-05,
+      "loss": 2.571,
+      "step": 262000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 2.5674,
+      "step": 263000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 2.5704,
+      "step": 264000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 2.5651,
+      "step": 265000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 2.566,
+      "step": 266000
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 5e-05,
+      "loss": 2.5664,
+      "step": 267000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 2.5699,
+      "step": 268000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 2.5628,
+      "step": 269000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 2.5494,
+      "step": 270000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 2.5604,
+      "step": 271000
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 5e-05,
+      "loss": 2.5563,
+      "step": 272000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 2.5598,
+      "step": 273000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 2.551,
+      "step": 274000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 2.5519,
+      "step": 275000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 2.5543,
+      "step": 276000
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 5e-05,
+      "loss": 2.5533,
+      "step": 277000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 2.552,
+      "step": 278000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 2.549,
+      "step": 279000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 2.5509,
+      "step": 280000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 2.5544,
+      "step": 281000
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 5e-05,
+      "loss": 2.5473,
+      "step": 282000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 2.543,
+      "step": 283000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 2.5403,
+      "step": 284000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 2.5488,
+      "step": 285000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 2.5361,
+      "step": 286000
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 5e-05,
+      "loss": 2.5385,
+      "step": 287000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 2.5422,
+      "step": 288000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 2.5363,
+      "step": 289000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 2.5344,
+      "step": 290000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 2.5315,
+      "step": 291000
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 5e-05,
+      "loss": 2.5345,
+      "step": 292000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 2.532,
+      "step": 293000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 2.5338,
+      "step": 294000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 2.5299,
+      "step": 295000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 2.5191,
+      "step": 296000
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 5e-05,
+      "loss": 2.5273,
+      "step": 297000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 2.5255,
+      "step": 298000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 2.5229,
+      "step": 299000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 2.5179,
+      "step": 300000
+    },
+    {
+      "epoch": 0.61,
+      "eval_accuracy": 0.559970362665269,
+      "eval_loss": 2.365234375,
+      "eval_runtime": 4356.0676,
+      "eval_samples_per_second": 146.024,
+      "eval_steps_per_second": 1.141,
+      "step": 300000
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 5e-05,
+      "loss": 2.5194,
+      "step": 301000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 2.5175,
+      "step": 302000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 2.5112,
+      "step": 303000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 2.5167,
+      "step": 304000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 2.5064,
+      "step": 305000
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 5e-05,
+      "loss": 2.5107,
+      "step": 306000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 2.5186,
+      "step": 307000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 2.5124,
+      "step": 308000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 2.5068,
+      "step": 309000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 2.5067,
+      "step": 310000
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 5e-05,
+      "loss": 2.5117,
+      "step": 311000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 2.5057,
+      "step": 312000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 2.5043,
+      "step": 313000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 2.5088,
+      "step": 314000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 2.5001,
+      "step": 315000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5e-05,
+      "loss": 2.5047,
+      "step": 316000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 2.502,
+      "step": 317000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 2.5013,
+      "step": 318000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 2.4957,
+      "step": 319000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 2.5014,
+      "step": 320000
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5e-05,
+      "loss": 2.4969,
+      "step": 321000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 2.4977,
+      "step": 322000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 2.4984,
+      "step": 323000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 2.4955,
+      "step": 324000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 2.4938,
+      "step": 325000
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5e-05,
+      "loss": 2.4898,
+      "step": 326000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 2.4916,
+      "step": 327000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 2.4895,
+      "step": 328000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 2.488,
+      "step": 329000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 2.4883,
+      "step": 330000
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5e-05,
+      "loss": 2.4857,
+      "step": 331000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 2.4836,
+      "step": 332000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 2.4799,
+      "step": 333000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 2.4858,
+      "step": 334000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 2.4765,
+      "step": 335000
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5e-05,
+      "loss": 2.4771,
+      "step": 336000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 2.4756,
+      "step": 337000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 2.4804,
+      "step": 338000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 2.472,
+      "step": 339000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 2.4687,
+      "step": 340000
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 5e-05,
+      "loss": 2.4754,
+      "step": 341000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 2.472,
+      "step": 342000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 2.4764,
+      "step": 343000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 2.4749,
+      "step": 344000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 2.4738,
+      "step": 345000
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 5e-05,
+      "loss": 2.4738,
+      "step": 346000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 2.4718,
+      "step": 347000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 2.4636,
+      "step": 348000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 2.4647,
+      "step": 349000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 2.4698,
+      "step": 350000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 5e-05,
+      "loss": 2.4628,
+      "step": 351000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 2.4639,
+      "step": 352000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 2.4678,
+      "step": 353000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 2.4631,
+      "step": 354000
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 5e-05,
+      "loss": 2.4603,
+      "step": 355000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 2.4574,
+      "step": 356000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 2.4646,
+      "step": 357000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 2.4569,
+      "step": 358000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 2.4532,
+      "step": 359000
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 5e-05,
+      "loss": 2.454,
+      "step": 360000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 2.4575,
+      "step": 361000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 2.4595,
+      "step": 362000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 2.4521,
+      "step": 363000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 2.4497,
+      "step": 364000
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 5e-05,
+      "loss": 2.4498,
+      "step": 365000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 2.4557,
+      "step": 366000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 2.4539,
+      "step": 367000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 2.4554,
+      "step": 368000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 2.4493,
+      "step": 369000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 2.4545,
+      "step": 370000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.4434,
+      "step": 371000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.4434,
+      "step": 372000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.4466,
+      "step": 373000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.4404,
+      "step": 374000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.4474,
+      "step": 375000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.448,
+      "step": 376000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.438,
+      "step": 377000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.4293,
+      "step": 378000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.4345,
+      "step": 379000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.4353,
+      "step": 380000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.4443,
+      "step": 381000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.4342,
+      "step": 382000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.4366,
+      "step": 383000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.4333,
+      "step": 384000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.4332,
+      "step": 385000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.4417,
+      "step": 386000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.4337,
+      "step": 387000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.4338,
+      "step": 388000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.4343,
+      "step": 389000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.4296,
+      "step": 390000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.4279,
+      "step": 391000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.4267,
+      "step": 392000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.4282,
+      "step": 393000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.423,
+      "step": 394000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.4308,
+      "step": 395000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.4268,
+      "step": 396000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.4241,
+      "step": 397000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.4235,
+      "step": 398000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.4261,
+      "step": 399000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.4277,
+      "step": 400000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.4179,
+      "step": 401000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.4187,
+      "step": 402000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.422,
+      "step": 403000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.4162,
+      "step": 404000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.4152,
+      "step": 405000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.4172,
+      "step": 406000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.4204,
+      "step": 407000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.4138,
+      "step": 408000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.4152,
+      "step": 409000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.4157,
+      "step": 410000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.4187,
+      "step": 411000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.4146,
+      "step": 412000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.4143,
+      "step": 413000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.4064,
+      "step": 414000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.4121,
+      "step": 415000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.4184,
+      "step": 416000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.4115,
+      "step": 417000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.4108,
+      "step": 418000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.4053,
+      "step": 419000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.4071,
+      "step": 420000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.4116,
+      "step": 421000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.4053,
+      "step": 422000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.4033,
+      "step": 423000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.4029,
+      "step": 424000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.399,
+      "step": 425000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3942,
+      "step": 426000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.4024,
+      "step": 427000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.4029,
+      "step": 428000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3979,
+      "step": 429000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.4071,
+      "step": 430000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.4013,
+      "step": 431000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3988,
+      "step": 432000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3961,
+      "step": 433000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.397,
+      "step": 434000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3924,
+      "step": 435000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3951,
+      "step": 436000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.394,
+      "step": 437000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.392,
+      "step": 438000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3939,
+      "step": 439000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3875,
+      "step": 440000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3946,
+      "step": 441000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3964,
+      "step": 442000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3913,
+      "step": 443000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3847,
+      "step": 444000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3848,
+      "step": 445000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3913,
+      "step": 446000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3868,
+      "step": 447000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3861,
+      "step": 448000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3807,
+      "step": 449000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3852,
+      "step": 450000
+    },
+    {
+      "epoch": 0.92,
+      "eval_accuracy": 0.5776840726507144,
+      "eval_loss": 2.23046875,
+      "eval_runtime": 4352.3033,
+      "eval_samples_per_second": 146.15,
+      "eval_steps_per_second": 1.142,
+      "step": 450000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3857,
+      "step": 451000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 452000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3847,
+      "step": 453000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3855,
+      "step": 454000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3818,
+      "step": 455000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3804,
+      "step": 456000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3855,
+      "step": 457000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3847,
+      "step": 458000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3857,
+      "step": 459000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3729,
+      "step": 460000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.377,
+      "step": 461000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3784,
+      "step": 462000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.374,
+      "step": 463000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.377,
+      "step": 464000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3847,
+      "step": 465000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3745,
+      "step": 466000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3809,
+      "step": 467000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.378,
+      "step": 468000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3776,
+      "step": 469000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3706,
+      "step": 470000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3707,
+      "step": 471000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3761,
+      "step": 472000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3766,
+      "step": 473000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3704,
+      "step": 474000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3698,
+      "step": 475000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.361,
+      "step": 476000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.37,
+      "step": 477000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3639,
+      "step": 478000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3656,
+      "step": 479000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3704,
+      "step": 480000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3593,
+      "step": 481000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3684,
+      "step": 482000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.363,
+      "step": 483000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3646,
+      "step": 484000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3638,
+      "step": 485000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3579,
+      "step": 486000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3678,
+      "step": 487000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3608,
+      "step": 488000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3639,
+      "step": 489000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.359,
+      "step": 490000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3676,
+      "step": 491000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.38,
+      "step": 492000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3867,
+      "step": 493000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3887,
+      "step": 494000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3917,
+      "step": 495000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.4049,
+      "step": 496000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3964,
+      "step": 497000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.4028,
+      "step": 498000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4004,
+      "step": 499000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4035,
+      "step": 500000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.397,
+      "step": 501000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.3992,
+      "step": 502000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4069,
+      "step": 503000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4047,
+      "step": 504000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4067,
+      "step": 505000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4047,
+      "step": 506000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.406,
+      "step": 507000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4089,
+      "step": 508000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4097,
+      "step": 509000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4079,
+      "step": 510000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4092,
+      "step": 511000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4071,
+      "step": 512000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4099,
+      "step": 513000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4075,
+      "step": 514000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4157,
+      "step": 515000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4169,
+      "step": 516000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4146,
+      "step": 517000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4118,
+      "step": 518000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4097,
+      "step": 519000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4072,
+      "step": 520000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4095,
+      "step": 521000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.411,
+      "step": 522000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4136,
+      "step": 523000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4114,
+      "step": 524000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4107,
+      "step": 525000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4148,
+      "step": 526000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4154,
+      "step": 527000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4142,
+      "step": 528000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4168,
+      "step": 529000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4195,
+      "step": 530000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4123,
+      "step": 531000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4143,
+      "step": 532000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4077,
+      "step": 533000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4118,
+      "step": 534000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4117,
+      "step": 535000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4128,
+      "step": 536000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4087,
+      "step": 537000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4105,
+      "step": 538000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4088,
+      "step": 539000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4101,
+      "step": 540000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4183,
+      "step": 541000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4111,
+      "step": 542000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4158,
+      "step": 543000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4133,
+      "step": 544000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4061,
+      "step": 545000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4076,
+      "step": 546000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4027,
+      "step": 547000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4116,
+      "step": 548000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4056,
+      "step": 549000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4049,
+      "step": 550000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4156,
+      "step": 551000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4069,
+      "step": 552000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.413,
+      "step": 553000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4145,
+      "step": 554000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4054,
+      "step": 555000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4105,
+      "step": 556000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.411,
+      "step": 557000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4102,
+      "step": 558000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4046,
+      "step": 559000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.418,
+      "step": 560000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4127,
+      "step": 561000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4092,
+      "step": 562000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4132,
+      "step": 563000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4044,
+      "step": 564000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4066,
+      "step": 565000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.408,
+      "step": 566000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4036,
+      "step": 567000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4081,
+      "step": 568000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4057,
+      "step": 569000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4054,
+      "step": 570000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4042,
+      "step": 571000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4097,
+      "step": 572000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.402,
+      "step": 573000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.406,
+      "step": 574000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4076,
+      "step": 575000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4037,
+      "step": 576000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4065,
+      "step": 577000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4057,
+      "step": 578000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4074,
+      "step": 579000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4093,
+      "step": 580000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4106,
+      "step": 581000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4098,
+      "step": 582000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4111,
+      "step": 583000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.407,
+      "step": 584000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4061,
+      "step": 585000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4093,
+      "step": 586000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4131,
+      "step": 587000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4063,
+      "step": 588000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.3983,
+      "step": 589000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.3993,
+      "step": 590000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4097,
+      "step": 591000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4116,
+      "step": 592000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4072,
+      "step": 593000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4063,
+      "step": 594000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4001,
+      "step": 595000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4022,
+      "step": 596000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4016,
+      "step": 597000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4054,
+      "step": 598000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4033,
+      "step": 599000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4014,
+      "step": 600000
+    },
+    {
+      "epoch": 1.22,
+      "eval_accuracy": 0.5762903057161931,
+      "eval_loss": 2.248046875,
+      "eval_runtime": 4358.0709,
+      "eval_samples_per_second": 145.957,
+      "eval_steps_per_second": 1.14,
+      "step": 600000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4013,
+      "step": 601000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4027,
+      "step": 602000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4046,
+      "step": 603000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4089,
+      "step": 604000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.402,
+      "step": 605000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4032,
+      "step": 606000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.3969,
+      "step": 607000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.3968,
+      "step": 608000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.3976,
+      "step": 609000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.405,
+      "step": 610000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.396,
+      "step": 611000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4035,
+      "step": 612000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4045,
+      "step": 613000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4028,
+      "step": 614000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4025,
+      "step": 615000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4006,
+      "step": 616000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4004,
+      "step": 617000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.3981,
+      "step": 618000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4053,
+      "step": 619000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.3995,
+      "step": 620000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.3994,
+      "step": 621000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.3977,
+      "step": 622000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4009,
+      "step": 623000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4015,
+      "step": 624000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.3978,
+      "step": 625000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.3966,
+      "step": 626000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.3992,
+      "step": 627000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.3999,
+      "step": 628000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.3962,
+      "step": 629000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.3965,
+      "step": 630000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.3983,
+      "step": 631000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.3979,
+      "step": 632000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.3962,
+      "step": 633000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.3931,
+      "step": 634000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.3965,
+      "step": 635000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.3975,
+      "step": 636000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.3923,
+      "step": 637000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.3994,
+      "step": 638000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.3969,
+      "step": 639000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.3968,
+      "step": 640000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.3943,
+      "step": 641000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.3919,
+      "step": 642000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.3971,
+      "step": 643000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.3995,
+      "step": 644000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.3909,
+      "step": 645000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4019,
+      "step": 646000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.3957,
+      "step": 647000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.3901,
+      "step": 648000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.3967,
+      "step": 649000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.3922,
+      "step": 650000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.3917,
+      "step": 651000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.3872,
+      "step": 652000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.3969,
+      "step": 653000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.3913,
+      "step": 654000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.3949,
+      "step": 655000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.3943,
+      "step": 656000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4,
+      "step": 657000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.3876,
+      "step": 658000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.3951,
+      "step": 659000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.393,
+      "step": 660000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3904,
+      "step": 661000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3896,
+      "step": 662000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3896,
+      "step": 663000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3906,
+      "step": 664000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3882,
+      "step": 665000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3878,
+      "step": 666000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 667000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3857,
+      "step": 668000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3902,
+      "step": 669000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3943,
+      "step": 670000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3948,
+      "step": 671000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3866,
+      "step": 672000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3854,
+      "step": 673000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3835,
+      "step": 674000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.387,
+      "step": 675000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3926,
+      "step": 676000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3857,
+      "step": 677000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3924,
+      "step": 678000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3847,
+      "step": 679000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3876,
+      "step": 680000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3831,
+      "step": 681000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3858,
+      "step": 682000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3895,
+      "step": 683000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3885,
+      "step": 684000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3884,
+      "step": 685000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 686000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.384,
+      "step": 687000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3894,
+      "step": 688000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 689000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3805,
+      "step": 690000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3866,
+      "step": 691000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3893,
+      "step": 692000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3863,
+      "step": 693000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 694000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3872,
+      "step": 695000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3819,
+      "step": 696000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3841,
+      "step": 697000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.383,
+      "step": 698000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3881,
+      "step": 699000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3789,
+      "step": 700000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3898,
+      "step": 701000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.387,
+      "step": 702000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3809,
+      "step": 703000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.381,
+      "step": 704000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3913,
+      "step": 705000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3808,
+      "step": 706000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.373,
+      "step": 707000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3858,
+      "step": 708000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3818,
+      "step": 709000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3791,
+      "step": 710000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3801,
+      "step": 711000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3827,
+      "step": 712000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3809,
+      "step": 713000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3788,
+      "step": 714000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3793,
+      "step": 715000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3795,
+      "step": 716000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3744,
+      "step": 717000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.383,
+      "step": 718000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3708,
+      "step": 719000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3771,
+      "step": 720000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3825,
+      "step": 721000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3769,
+      "step": 722000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3809,
+      "step": 723000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3814,
+      "step": 724000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3737,
+      "step": 725000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3754,
+      "step": 726000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3774,
+      "step": 727000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.38,
+      "step": 728000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3769,
+      "step": 729000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3783,
+      "step": 730000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3766,
+      "step": 731000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3771,
+      "step": 732000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3799,
+      "step": 733000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3794,
+      "step": 734000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3684,
+      "step": 735000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3789,
+      "step": 736000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3809,
+      "step": 737000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3753,
+      "step": 738000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3762,
+      "step": 739000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3747,
+      "step": 740000
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 5e-05,
+      "loss": 2.3634,
+      "step": 741000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3763,
+      "step": 742000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3777,
+      "step": 743000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3716,
+      "step": 744000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3731,
+      "step": 745000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3762,
+      "step": 746000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3692,
+      "step": 747000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3725,
+      "step": 748000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3744,
+      "step": 749000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3687,
+      "step": 750000
+    },
+    {
+      "epoch": 0.76,
+      "eval_accuracy": 0.5800586539733796,
+      "eval_loss": 2.21484375,
+      "eval_runtime": 4527.6466,
+      "eval_samples_per_second": 140.49,
+      "eval_steps_per_second": 2.195,
+      "step": 750000
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5e-05,
+      "loss": 2.3717,
+      "step": 751000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3724,
+      "step": 752000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3708,
+      "step": 753000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3687,
+      "step": 754000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.369,
+      "step": 755000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3635,
+      "step": 756000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3656,
+      "step": 757000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3692,
+      "step": 758000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3686,
+      "step": 759000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3694,
+      "step": 760000
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5e-05,
+      "loss": 2.3721,
+      "step": 761000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3678,
+      "step": 762000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3692,
+      "step": 763000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3711,
+      "step": 764000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3741,
+      "step": 765000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3693,
+      "step": 766000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3775,
+      "step": 767000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3654,
+      "step": 768000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.3756,
+      "step": 769000
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5e-05,
+      "loss": 2.363,
+      "step": 770000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.374,
+      "step": 771000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3705,
+      "step": 772000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3719,
+      "step": 773000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3705,
+      "step": 774000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3702,
+      "step": 775000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3671,
+      "step": 776000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3707,
+      "step": 777000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.372,
+      "step": 778000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3781,
+      "step": 779000
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 5e-05,
+      "loss": 2.3755,
+      "step": 780000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3614,
+      "step": 781000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3564,
+      "step": 782000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3594,
+      "step": 783000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3725,
+      "step": 784000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3764,
+      "step": 785000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3659,
+      "step": 786000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3611,
+      "step": 787000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.363,
+      "step": 788000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3702,
+      "step": 789000
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 5e-05,
+      "loss": 2.3609,
+      "step": 790000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3677,
+      "step": 791000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3635,
+      "step": 792000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3654,
+      "step": 793000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3711,
+      "step": 794000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3589,
+      "step": 795000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3601,
+      "step": 796000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.363,
+      "step": 797000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3693,
+      "step": 798000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.3613,
+      "step": 799000
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 5e-05,
+      "loss": 2.363,
+      "step": 800000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3636,
+      "step": 801000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3649,
+      "step": 802000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3687,
+      "step": 803000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3635,
+      "step": 804000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3605,
+      "step": 805000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3718,
+      "step": 806000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3472,
+      "step": 807000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3656,
+      "step": 808000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3652,
+      "step": 809000
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 5e-05,
+      "loss": 2.3637,
+      "step": 810000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3647,
+      "step": 811000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3579,
+      "step": 812000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3628,
+      "step": 813000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3593,
+      "step": 814000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3517,
+      "step": 815000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3665,
+      "step": 816000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3578,
+      "step": 817000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3612,
+      "step": 818000
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 5e-05,
+      "loss": 2.3625,
+      "step": 819000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3614,
+      "step": 820000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3685,
+      "step": 821000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3642,
+      "step": 822000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3645,
+      "step": 823000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3541,
+      "step": 824000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3575,
+      "step": 825000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3511,
+      "step": 826000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3555,
+      "step": 827000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3575,
+      "step": 828000
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 5e-05,
+      "loss": 2.3585,
+      "step": 829000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3639,
+      "step": 830000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3656,
+      "step": 831000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3595,
+      "step": 832000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.359,
+      "step": 833000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3586,
+      "step": 834000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3516,
+      "step": 835000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3609,
+      "step": 836000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.351,
+      "step": 837000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3596,
+      "step": 838000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 5e-05,
+      "loss": 2.3511,
+      "step": 839000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3579,
+      "step": 840000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3591,
+      "step": 841000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3498,
+      "step": 842000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3606,
+      "step": 843000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3504,
+      "step": 844000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3442,
+      "step": 845000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3553,
+      "step": 846000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.351,
+      "step": 847000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3511,
+      "step": 848000
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 5e-05,
+      "loss": 2.3573,
+      "step": 849000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3529,
+      "step": 850000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.352,
+      "step": 851000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3507,
+      "step": 852000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3595,
+      "step": 853000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.353,
+      "step": 854000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3444,
+      "step": 855000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3568,
+      "step": 856000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3547,
+      "step": 857000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.362,
+      "step": 858000
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 5e-05,
+      "loss": 2.3635,
+      "step": 859000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3548,
+      "step": 860000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3522,
+      "step": 861000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.355,
+      "step": 862000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3583,
+      "step": 863000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3446,
+      "step": 864000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3536,
+      "step": 865000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3507,
+      "step": 866000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3493,
+      "step": 867000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3568,
+      "step": 868000
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 5e-05,
+      "loss": 2.3498,
+      "step": 869000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3537,
+      "step": 870000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3622,
+      "step": 871000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3551,
+      "step": 872000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.353,
+      "step": 873000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.358,
+      "step": 874000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.359,
+      "step": 875000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3573,
+      "step": 876000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3601,
+      "step": 877000
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5e-05,
+      "loss": 2.3546,
+      "step": 878000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3502,
+      "step": 879000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3465,
+      "step": 880000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3556,
+      "step": 881000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3415,
+      "step": 882000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3538,
+      "step": 883000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.356,
+      "step": 884000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3469,
+      "step": 885000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3536,
+      "step": 886000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3509,
+      "step": 887000
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5e-05,
+      "loss": 2.3441,
+      "step": 888000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.344,
+      "step": 889000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3485,
+      "step": 890000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3453,
+      "step": 891000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3492,
+      "step": 892000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3476,
+      "step": 893000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3411,
+      "step": 894000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3497,
+      "step": 895000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3413,
+      "step": 896000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3442,
+      "step": 897000
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 5e-05,
+      "loss": 2.3428,
+      "step": 898000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3419,
+      "step": 899000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3458,
+      "step": 900000
+    },
+    {
+      "epoch": 0.92,
+      "eval_accuracy": 0.5835522342641007,
+      "eval_loss": 2.19140625,
+      "eval_runtime": 4531.41,
+      "eval_samples_per_second": 140.373,
+      "eval_steps_per_second": 2.193,
+      "step": 900000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3477,
+      "step": 901000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.345,
+      "step": 902000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3438,
+      "step": 903000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3572,
+      "step": 904000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3532,
+      "step": 905000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3493,
+      "step": 906000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3419,
+      "step": 907000
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5e-05,
+      "loss": 2.3499,
+      "step": 908000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3388,
+      "step": 909000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.341,
+      "step": 910000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3441,
+      "step": 911000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3413,
+      "step": 912000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3448,
+      "step": 913000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3382,
+      "step": 914000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3463,
+      "step": 915000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3373,
+      "step": 916000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3429,
+      "step": 917000
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5e-05,
+      "loss": 2.3547,
+      "step": 918000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3407,
+      "step": 919000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3362,
+      "step": 920000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.341,
+      "step": 921000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3326,
+      "step": 922000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3431,
+      "step": 923000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3392,
+      "step": 924000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3372,
+      "step": 925000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3488,
+      "step": 926000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3382,
+      "step": 927000
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 5e-05,
+      "loss": 2.3445,
+      "step": 928000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3491,
+      "step": 929000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3487,
+      "step": 930000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3448,
+      "step": 931000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3407,
+      "step": 932000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.339,
+      "step": 933000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3402,
+      "step": 934000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3441,
+      "step": 935000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3357,
+      "step": 936000
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 5e-05,
+      "loss": 2.3493,
+      "step": 937000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3436,
+      "step": 938000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3453,
+      "step": 939000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3427,
+      "step": 940000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3394,
+      "step": 941000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3384,
+      "step": 942000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3375,
+      "step": 943000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3411,
+      "step": 944000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3375,
+      "step": 945000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.33,
+      "step": 946000
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 5e-05,
+      "loss": 2.3397,
+      "step": 947000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3365,
+      "step": 948000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3365,
+      "step": 949000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3355,
+      "step": 950000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3296,
+      "step": 951000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3364,
+      "step": 952000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3384,
+      "step": 953000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3443,
+      "step": 954000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3459,
+      "step": 955000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3342,
+      "step": 956000
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5e-05,
+      "loss": 2.3411,
+      "step": 957000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3334,
+      "step": 958000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3428,
+      "step": 959000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3408,
+      "step": 960000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3349,
+      "step": 961000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3406,
+      "step": 962000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3305,
+      "step": 963000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3381,
+      "step": 964000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3392,
+      "step": 965000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3371,
+      "step": 966000
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5e-05,
+      "loss": 2.3259,
+      "step": 967000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3422,
+      "step": 968000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3402,
+      "step": 969000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3354,
+      "step": 970000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3368,
+      "step": 971000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3369,
+      "step": 972000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.33,
+      "step": 973000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3311,
+      "step": 974000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3476,
+      "step": 975000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.3384,
+      "step": 976000
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5e-05,
+      "loss": 2.331,
+      "step": 977000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.331,
+      "step": 978000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3305,
+      "step": 979000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.341,
+      "step": 980000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3358,
+      "step": 981000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3383,
+      "step": 982000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3546,
+      "step": 983000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3664,
+      "step": 984000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3786,
+      "step": 985000
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5e-05,
+      "loss": 2.3733,
+      "step": 986000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.385,
+      "step": 987000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3884,
+      "step": 988000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3883,
+      "step": 989000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3809,
+      "step": 990000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3811,
+      "step": 991000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.395,
+      "step": 992000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3891,
+      "step": 993000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3968,
+      "step": 994000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.4015,
+      "step": 995000
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 5e-05,
+      "loss": 2.3996,
+      "step": 996000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4019,
+      "step": 997000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.3913,
+      "step": 998000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4018,
+      "step": 999000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4003,
+      "step": 1000000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4019,
+      "step": 1001000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4067,
+      "step": 1002000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4113,
+      "step": 1003000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.407,
+      "step": 1004000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.3967,
+      "step": 1005000
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 5e-05,
+      "loss": 2.4192,
+      "step": 1006000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4173,
+      "step": 1007000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4183,
+      "step": 1008000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4283,
+      "step": 1009000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4155,
+      "step": 1010000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4143,
+      "step": 1011000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4163,
+      "step": 1012000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4094,
+      "step": 1013000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4105,
+      "step": 1014000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4184,
+      "step": 1015000
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 5e-05,
+      "loss": 2.4119,
+      "step": 1016000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4216,
+      "step": 1017000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4201,
+      "step": 1018000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.414,
+      "step": 1019000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4188,
+      "step": 1020000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4212,
+      "step": 1021000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4234,
+      "step": 1022000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4205,
+      "step": 1023000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4175,
+      "step": 1024000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4264,
+      "step": 1025000
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 5e-05,
+      "loss": 2.4271,
+      "step": 1026000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4244,
+      "step": 1027000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4271,
+      "step": 1028000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4229,
+      "step": 1029000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4247,
+      "step": 1030000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4273,
+      "step": 1031000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4278,
+      "step": 1032000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4232,
+      "step": 1033000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4277,
+      "step": 1034000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4329,
+      "step": 1035000
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 5e-05,
+      "loss": 2.4349,
+      "step": 1036000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4195,
+      "step": 1037000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4233,
+      "step": 1038000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4275,
+      "step": 1039000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4348,
+      "step": 1040000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4397,
+      "step": 1041000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4249,
+      "step": 1042000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4282,
+      "step": 1043000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4254,
+      "step": 1044000
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 5e-05,
+      "loss": 2.4235,
+      "step": 1045000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4315,
+      "step": 1046000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.426,
+      "step": 1047000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4344,
+      "step": 1048000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4269,
+      "step": 1049000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4338,
+      "step": 1050000
+    },
+    {
+      "epoch": 1.07,
+      "eval_accuracy": 0.5744690190170737,
+      "eval_loss": 2.263671875,
+      "eval_runtime": 4537.1601,
+      "eval_samples_per_second": 140.195,
+      "eval_steps_per_second": 2.191,
+      "step": 1050000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.433,
+      "step": 1051000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4197,
+      "step": 1052000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4293,
+      "step": 1053000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4331,
+      "step": 1054000
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 5e-05,
+      "loss": 2.4348,
+      "step": 1055000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4413,
+      "step": 1056000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4263,
+      "step": 1057000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4423,
+      "step": 1058000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4382,
+      "step": 1059000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4396,
+      "step": 1060000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4401,
+      "step": 1061000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4363,
+      "step": 1062000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4345,
+      "step": 1063000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4344,
+      "step": 1064000
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 5e-05,
+      "loss": 2.4401,
+      "step": 1065000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4341,
+      "step": 1066000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4318,
+      "step": 1067000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4309,
+      "step": 1068000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4464,
+      "step": 1069000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4363,
+      "step": 1070000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4425,
+      "step": 1071000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4393,
+      "step": 1072000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4332,
+      "step": 1073000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4361,
+      "step": 1074000
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 5e-05,
+      "loss": 2.4362,
+      "step": 1075000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4318,
+      "step": 1076000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4382,
+      "step": 1077000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4291,
+      "step": 1078000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.445,
+      "step": 1079000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4356,
+      "step": 1080000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4381,
+      "step": 1081000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.442,
+      "step": 1082000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4386,
+      "step": 1083000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4339,
+      "step": 1084000
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 5e-05,
+      "loss": 2.4461,
+      "step": 1085000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4401,
+      "step": 1086000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4379,
+      "step": 1087000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4378,
+      "step": 1088000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4359,
+      "step": 1089000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4364,
+      "step": 1090000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4416,
+      "step": 1091000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4464,
+      "step": 1092000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4335,
+      "step": 1093000
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 5e-05,
+      "loss": 2.4319,
+      "step": 1094000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4364,
+      "step": 1095000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4337,
+      "step": 1096000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.436,
+      "step": 1097000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4354,
+      "step": 1098000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4483,
+      "step": 1099000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4407,
+      "step": 1100000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4465,
+      "step": 1101000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4479,
+      "step": 1102000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4437,
+      "step": 1103000
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 5e-05,
+      "loss": 2.4388,
+      "step": 1104000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4465,
+      "step": 1105000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.448,
+      "step": 1106000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4379,
+      "step": 1107000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.445,
+      "step": 1108000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4388,
+      "step": 1109000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.436,
+      "step": 1110000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4404,
+      "step": 1111000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4369,
+      "step": 1112000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4421,
+      "step": 1113000
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 5e-05,
+      "loss": 2.4451,
+      "step": 1114000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4462,
+      "step": 1115000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.44,
+      "step": 1116000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4405,
+      "step": 1117000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4447,
+      "step": 1118000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4421,
+      "step": 1119000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4524,
+      "step": 1120000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4429,
+      "step": 1121000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4386,
+      "step": 1122000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.4475,
+      "step": 1123000
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 5e-05,
+      "loss": 2.445,
+      "step": 1124000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4517,
+      "step": 1125000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.446,
+      "step": 1126000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4447,
+      "step": 1127000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4461,
+      "step": 1128000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4457,
+      "step": 1129000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4386,
+      "step": 1130000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4454,
+      "step": 1131000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.454,
+      "step": 1132000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4392,
+      "step": 1133000
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 5e-05,
+      "loss": 2.4389,
+      "step": 1134000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4478,
+      "step": 1135000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4402,
+      "step": 1136000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4387,
+      "step": 1137000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4421,
+      "step": 1138000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4397,
+      "step": 1139000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4474,
+      "step": 1140000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4502,
+      "step": 1141000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4471,
+      "step": 1142000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4368,
+      "step": 1143000
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 5e-05,
+      "loss": 2.4472,
+      "step": 1144000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4434,
+      "step": 1145000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4479,
+      "step": 1146000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4504,
+      "step": 1147000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4485,
+      "step": 1148000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4407,
+      "step": 1149000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4487,
+      "step": 1150000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4456,
+      "step": 1151000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4463,
+      "step": 1152000
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 5e-05,
+      "loss": 2.4359,
+      "step": 1153000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4479,
+      "step": 1154000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4425,
+      "step": 1155000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4485,
+      "step": 1156000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4507,
+      "step": 1157000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.449,
+      "step": 1158000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4403,
+      "step": 1159000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4432,
+      "step": 1160000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4453,
+      "step": 1161000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4561,
+      "step": 1162000
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 5e-05,
+      "loss": 2.4474,
+      "step": 1163000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4475,
+      "step": 1164000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4445,
+      "step": 1165000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4483,
+      "step": 1166000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4439,
+      "step": 1167000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4436,
+      "step": 1168000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4485,
+      "step": 1169000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.451,
+      "step": 1170000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4502,
+      "step": 1171000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4446,
+      "step": 1172000
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 5e-05,
+      "loss": 2.4529,
+      "step": 1173000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.45,
+      "step": 1174000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4436,
+      "step": 1175000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4536,
+      "step": 1176000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4422,
+      "step": 1177000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4537,
+      "step": 1178000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.446,
+      "step": 1179000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4463,
+      "step": 1180000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4491,
+      "step": 1181000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4455,
+      "step": 1182000
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 5e-05,
+      "loss": 2.4466,
+      "step": 1183000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4589,
+      "step": 1184000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4432,
+      "step": 1185000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4566,
+      "step": 1186000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4535,
+      "step": 1187000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4558,
+      "step": 1188000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4543,
+      "step": 1189000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4473,
+      "step": 1190000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4506,
+      "step": 1191000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4444,
+      "step": 1192000
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 5e-05,
+      "loss": 2.4341,
+      "step": 1193000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4428,
+      "step": 1194000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4514,
+      "step": 1195000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4514,
+      "step": 1196000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4538,
+      "step": 1197000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4422,
+      "step": 1198000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4455,
+      "step": 1199000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.448,
+      "step": 1200000
+    },
+    {
+      "epoch": 1.22,
+      "eval_accuracy": 0.5731948419940126,
+      "eval_loss": 2.279296875,
+      "eval_runtime": 4533.1358,
+      "eval_samples_per_second": 140.32,
+      "eval_steps_per_second": 2.193,
+      "step": 1200000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4536,
+      "step": 1201000
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 5e-05,
+      "loss": 2.4473,
+      "step": 1202000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4483,
+      "step": 1203000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4516,
+      "step": 1204000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4508,
+      "step": 1205000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4462,
+      "step": 1206000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4443,
+      "step": 1207000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4477,
+      "step": 1208000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4552,
+      "step": 1209000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4441,
+      "step": 1210000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4366,
+      "step": 1211000
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 5e-05,
+      "loss": 2.4502,
+      "step": 1212000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4434,
+      "step": 1213000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4413,
+      "step": 1214000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4485,
+      "step": 1215000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4448,
+      "step": 1216000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4434,
+      "step": 1217000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4505,
+      "step": 1218000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4532,
+      "step": 1219000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4433,
+      "step": 1220000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.444,
+      "step": 1221000
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 5e-05,
+      "loss": 2.4528,
+      "step": 1222000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4537,
+      "step": 1223000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4518,
+      "step": 1224000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.453,
+      "step": 1225000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4527,
+      "step": 1226000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.45,
+      "step": 1227000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4515,
+      "step": 1228000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.451,
+      "step": 1229000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4475,
+      "step": 1230000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4525,
+      "step": 1231000
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 5e-05,
+      "loss": 2.4522,
+      "step": 1232000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4496,
+      "step": 1233000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4548,
+      "step": 1234000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4486,
+      "step": 1235000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4499,
+      "step": 1236000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4609,
+      "step": 1237000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4543,
+      "step": 1238000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4587,
+      "step": 1239000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4502,
+      "step": 1240000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4547,
+      "step": 1241000
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 5e-05,
+      "loss": 2.4427,
+      "step": 1242000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4481,
+      "step": 1243000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4528,
+      "step": 1244000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4512,
+      "step": 1245000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4537,
+      "step": 1246000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4622,
+      "step": 1247000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4572,
+      "step": 1248000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4423,
+      "step": 1249000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4502,
+      "step": 1250000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4491,
+      "step": 1251000
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 5e-05,
+      "loss": 2.4474,
+      "step": 1252000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4418,
+      "step": 1253000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4471,
+      "step": 1254000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4493,
+      "step": 1255000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4531,
+      "step": 1256000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4544,
+      "step": 1257000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4502,
+      "step": 1258000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4457,
+      "step": 1259000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4434,
+      "step": 1260000
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 5e-05,
+      "loss": 2.4529,
+      "step": 1261000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4481,
+      "step": 1262000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4427,
+      "step": 1263000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4508,
+      "step": 1264000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4528,
+      "step": 1265000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4481,
+      "step": 1266000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4456,
+      "step": 1267000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4483,
+      "step": 1268000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4458,
+      "step": 1269000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4463,
+      "step": 1270000
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 5e-05,
+      "loss": 2.4517,
+      "step": 1271000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4485,
+      "step": 1272000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4547,
+      "step": 1273000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4532,
+      "step": 1274000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4449,
+      "step": 1275000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4499,
+      "step": 1276000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4534,
+      "step": 1277000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4556,
+      "step": 1278000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.451,
+      "step": 1279000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4432,
+      "step": 1280000
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 5e-05,
+      "loss": 2.4562,
+      "step": 1281000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4352,
+      "step": 1282000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4479,
+      "step": 1283000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4497,
+      "step": 1284000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4546,
+      "step": 1285000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4498,
+      "step": 1286000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4524,
+      "step": 1287000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4556,
+      "step": 1288000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4564,
+      "step": 1289000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4509,
+      "step": 1290000
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 5e-05,
+      "loss": 2.4478,
+      "step": 1291000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4542,
+      "step": 1292000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4508,
+      "step": 1293000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4499,
+      "step": 1294000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.443,
+      "step": 1295000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4505,
+      "step": 1296000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4512,
+      "step": 1297000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.455,
+      "step": 1298000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4506,
+      "step": 1299000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4508,
+      "step": 1300000
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 5e-05,
+      "loss": 2.4427,
+      "step": 1301000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4479,
+      "step": 1302000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4501,
+      "step": 1303000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4427,
+      "step": 1304000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4547,
+      "step": 1305000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4496,
+      "step": 1306000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4526,
+      "step": 1307000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4464,
+      "step": 1308000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4545,
+      "step": 1309000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.4591,
+      "step": 1310000
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 5e-05,
+      "loss": 2.448,
+      "step": 1311000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4522,
+      "step": 1312000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4539,
+      "step": 1313000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4439,
+      "step": 1314000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4455,
+      "step": 1315000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4492,
+      "step": 1316000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4419,
+      "step": 1317000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4524,
+      "step": 1318000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4448,
+      "step": 1319000
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 5e-05,
+      "loss": 2.4492,
+      "step": 1320000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.4325,
+      "step": 1321000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.4132,
+      "step": 1322000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.4056,
+      "step": 1323000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.4033,
+      "step": 1324000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3908,
+      "step": 1325000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3877,
+      "step": 1326000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3788,
+      "step": 1327000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3885,
+      "step": 1328000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3796,
+      "step": 1329000
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 5e-05,
+      "loss": 2.3758,
+      "step": 1330000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3784,
+      "step": 1331000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3744,
+      "step": 1332000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3714,
+      "step": 1333000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.37,
+      "step": 1334000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3703,
+      "step": 1335000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3699,
+      "step": 1336000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3762,
+      "step": 1337000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3686,
+      "step": 1338000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.367,
+      "step": 1339000
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 5e-05,
+      "loss": 2.3674,
+      "step": 1340000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3608,
+      "step": 1341000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3576,
+      "step": 1342000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3544,
+      "step": 1343000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3598,
+      "step": 1344000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3538,
+      "step": 1345000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3541,
+      "step": 1346000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3516,
+      "step": 1347000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3544,
+      "step": 1348000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3611,
+      "step": 1349000
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 5e-05,
+      "loss": 2.3495,
+      "step": 1350000
+    },
+    {
+      "epoch": 1.37,
+      "eval_accuracy": 0.5847132515714882,
+      "eval_loss": 2.189453125,
+      "eval_runtime": 4583.3705,
+      "eval_samples_per_second": 138.782,
+      "eval_steps_per_second": 2.168,
+      "step": 1350000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3507,
+      "step": 1351000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3554,
+      "step": 1352000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.353,
+      "step": 1353000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3447,
+      "step": 1354000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3387,
+      "step": 1355000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3462,
+      "step": 1356000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3518,
+      "step": 1357000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3423,
+      "step": 1358000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.344,
+      "step": 1359000
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 5e-05,
+      "loss": 2.3389,
+      "step": 1360000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3394,
+      "step": 1361000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3378,
+      "step": 1362000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3493,
+      "step": 1363000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3445,
+      "step": 1364000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3373,
+      "step": 1365000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3373,
+      "step": 1366000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3432,
+      "step": 1367000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3381,
+      "step": 1368000
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 5e-05,
+      "loss": 2.3416,
+      "step": 1369000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3427,
+      "step": 1370000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3409,
+      "step": 1371000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3388,
+      "step": 1372000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3325,
+      "step": 1373000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3417,
+      "step": 1374000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3278,
+      "step": 1375000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.34,
+      "step": 1376000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3412,
+      "step": 1377000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3264,
+      "step": 1378000
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 5e-05,
+      "loss": 2.3228,
+      "step": 1379000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3236,
+      "step": 1380000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3253,
+      "step": 1381000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3373,
+      "step": 1382000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3336,
+      "step": 1383000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3261,
+      "step": 1384000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3319,
+      "step": 1385000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3242,
+      "step": 1386000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3269,
+      "step": 1387000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3278,
+      "step": 1388000
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 5e-05,
+      "loss": 2.3316,
+      "step": 1389000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3355,
+      "step": 1390000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3319,
+      "step": 1391000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3293,
+      "step": 1392000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3288,
+      "step": 1393000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3274,
+      "step": 1394000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3314,
+      "step": 1395000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3186,
+      "step": 1396000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3227,
+      "step": 1397000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.322,
+      "step": 1398000
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 5e-05,
+      "loss": 2.3138,
+      "step": 1399000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3208,
+      "step": 1400000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3294,
+      "step": 1401000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3162,
+      "step": 1402000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3337,
+      "step": 1403000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3245,
+      "step": 1404000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3247,
+      "step": 1405000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3232,
+      "step": 1406000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3227,
+      "step": 1407000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3232,
+      "step": 1408000
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 5e-05,
+      "loss": 2.3197,
+      "step": 1409000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3266,
+      "step": 1410000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3167,
+      "step": 1411000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3262,
+      "step": 1412000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3147,
+      "step": 1413000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3207,
+      "step": 1414000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3192,
+      "step": 1415000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3305,
+      "step": 1416000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3246,
+      "step": 1417000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3256,
+      "step": 1418000
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 5e-05,
+      "loss": 2.3168,
+      "step": 1419000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3142,
+      "step": 1420000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3165,
+      "step": 1421000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3179,
+      "step": 1422000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3056,
+      "step": 1423000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3195,
+      "step": 1424000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3149,
+      "step": 1425000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3111,
+      "step": 1426000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3164,
+      "step": 1427000
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 5e-05,
+      "loss": 2.3161,
+      "step": 1428000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3187,
+      "step": 1429000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3071,
+      "step": 1430000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3068,
+      "step": 1431000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.315,
+      "step": 1432000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3121,
+      "step": 1433000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3156,
+      "step": 1434000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3147,
+      "step": 1435000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3135,
+      "step": 1436000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.321,
+      "step": 1437000
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 5e-05,
+      "loss": 2.3172,
+      "step": 1438000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3135,
+      "step": 1439000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3141,
+      "step": 1440000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3186,
+      "step": 1441000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3118,
+      "step": 1442000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.315,
+      "step": 1443000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3212,
+      "step": 1444000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3052,
+      "step": 1445000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3075,
+      "step": 1446000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3156,
+      "step": 1447000
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 5e-05,
+      "loss": 2.3035,
+      "step": 1448000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3119,
+      "step": 1449000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3113,
+      "step": 1450000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3092,
+      "step": 1451000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3066,
+      "step": 1452000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.2963,
+      "step": 1453000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3148,
+      "step": 1454000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3098,
+      "step": 1455000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3118,
+      "step": 1456000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3153,
+      "step": 1457000
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 5e-05,
+      "loss": 2.3116,
+      "step": 1458000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3158,
+      "step": 1459000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.312,
+      "step": 1460000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3139,
+      "step": 1461000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3185,
+      "step": 1462000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3024,
+      "step": 1463000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3127,
+      "step": 1464000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.3058,
+      "step": 1465000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.317,
+      "step": 1466000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.2993,
+      "step": 1467000
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 5e-05,
+      "loss": 2.312,
+      "step": 1468000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3146,
+      "step": 1469000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3048,
+      "step": 1470000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3039,
+      "step": 1471000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3132,
+      "step": 1472000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3042,
+      "step": 1473000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3068,
+      "step": 1474000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3051,
+      "step": 1475000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.2953,
+      "step": 1476000
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 5e-05,
+      "loss": 2.3114,
+      "step": 1477000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3066,
+      "step": 1478000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3054,
+      "step": 1479000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3057,
+      "step": 1480000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.2906,
+      "step": 1481000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3048,
+      "step": 1482000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3021,
+      "step": 1483000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.2999,
+      "step": 1484000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3112,
+      "step": 1485000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3138,
+      "step": 1486000
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 5e-05,
+      "loss": 2.3054,
+      "step": 1487000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.3021,
+      "step": 1488000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.3084,
+      "step": 1489000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.2948,
+      "step": 1490000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.2951,
+      "step": 1491000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.3102,
+      "step": 1492000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.3027,
+      "step": 1493000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.2981,
+      "step": 1494000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.3074,
+      "step": 1495000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.2985,
+      "step": 1496000
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 5e-05,
+      "loss": 2.3069,
+      "step": 1497000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.2977,
+      "step": 1498000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.3008,
+      "step": 1499000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.2959,
+      "step": 1500000
+    },
+    {
+      "epoch": 1.53,
+      "eval_accuracy": 0.5908897665779291,
+      "eval_loss": 2.142578125,
+      "eval_runtime": 4555.5415,
+      "eval_samples_per_second": 139.63,
+      "eval_steps_per_second": 2.182,
+      "step": 1500000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.287,
+      "step": 1501000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.3021,
+      "step": 1502000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.2996,
+      "step": 1503000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.2913,
+      "step": 1504000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.3057,
+      "step": 1505000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.2943,
+      "step": 1506000
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 5e-05,
+      "loss": 2.3042,
+      "step": 1507000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.2943,
+      "step": 1508000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.2872,
+      "step": 1509000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.3081,
+      "step": 1510000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.3018,
+      "step": 1511000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.2979,
+      "step": 1512000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.295,
+      "step": 1513000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.2926,
+      "step": 1514000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.2908,
+      "step": 1515000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.3023,
+      "step": 1516000
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 5e-05,
+      "loss": 2.299,
+      "step": 1517000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2902,
+      "step": 1518000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.3034,
+      "step": 1519000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2973,
+      "step": 1520000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2937,
+      "step": 1521000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.296,
+      "step": 1522000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2924,
+      "step": 1523000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2941,
+      "step": 1524000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2903,
+      "step": 1525000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.299,
+      "step": 1526000
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 5e-05,
+      "loss": 2.2967,
+      "step": 1527000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.3052,
+      "step": 1528000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2998,
+      "step": 1529000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2943,
+      "step": 1530000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2938,
+      "step": 1531000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2934,
+      "step": 1532000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2964,
+      "step": 1533000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2961,
+      "step": 1534000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.3046,
+      "step": 1535000
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 5e-05,
+      "loss": 2.2901,
+      "step": 1536000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2947,
+      "step": 1537000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2914,
+      "step": 1538000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2977,
+      "step": 1539000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2945,
+      "step": 1540000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.302,
+      "step": 1541000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.3021,
+      "step": 1542000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2884,
+      "step": 1543000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2904,
+      "step": 1544000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2868,
+      "step": 1545000
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 5e-05,
+      "loss": 2.2917,
+      "step": 1546000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2833,
+      "step": 1547000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2923,
+      "step": 1548000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2978,
+      "step": 1549000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2921,
+      "step": 1550000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2945,
+      "step": 1551000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2815,
+      "step": 1552000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2857,
+      "step": 1553000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.289,
+      "step": 1554000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.2891,
+      "step": 1555000
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 5e-05,
+      "loss": 2.288,
+      "step": 1556000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2973,
+      "step": 1557000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2944,
+      "step": 1558000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2885,
+      "step": 1559000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2876,
+      "step": 1560000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.3017,
+      "step": 1561000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2916,
+      "step": 1562000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2881,
+      "step": 1563000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.28,
+      "step": 1564000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2944,
+      "step": 1565000
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 5e-05,
+      "loss": 2.2932,
+      "step": 1566000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2861,
+      "step": 1567000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2937,
+      "step": 1568000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2806,
+      "step": 1569000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2875,
+      "step": 1570000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2932,
+      "step": 1571000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2768,
+      "step": 1572000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2988,
+      "step": 1573000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2871,
+      "step": 1574000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2954,
+      "step": 1575000
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 5e-05,
+      "loss": 2.2979,
+      "step": 1576000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2966,
+      "step": 1577000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2931,
+      "step": 1578000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2885,
+      "step": 1579000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2843,
+      "step": 1580000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2924,
+      "step": 1581000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2885,
+      "step": 1582000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.283,
+      "step": 1583000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2894,
+      "step": 1584000
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 5e-05,
+      "loss": 2.2874,
+      "step": 1585000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2918,
+      "step": 1586000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.291,
+      "step": 1587000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2843,
+      "step": 1588000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2826,
+      "step": 1589000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2907,
+      "step": 1590000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.288,
+      "step": 1591000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2805,
+      "step": 1592000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2888,
+      "step": 1593000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2864,
+      "step": 1594000
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 5e-05,
+      "loss": 2.2861,
+      "step": 1595000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2837,
+      "step": 1596000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2849,
+      "step": 1597000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2866,
+      "step": 1598000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2842,
+      "step": 1599000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2848,
+      "step": 1600000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2813,
+      "step": 1601000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.288,
+      "step": 1602000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.282,
+      "step": 1603000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2938,
+      "step": 1604000
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 5e-05,
+      "loss": 2.2887,
+      "step": 1605000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2803,
+      "step": 1606000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2835,
+      "step": 1607000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2786,
+      "step": 1608000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2757,
+      "step": 1609000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2791,
+      "step": 1610000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.287,
+      "step": 1611000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2795,
+      "step": 1612000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2832,
+      "step": 1613000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2894,
+      "step": 1614000
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 5e-05,
+      "loss": 2.2815,
+      "step": 1615000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2759,
+      "step": 1616000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.283,
+      "step": 1617000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2862,
+      "step": 1618000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2804,
+      "step": 1619000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2825,
+      "step": 1620000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2738,
+      "step": 1621000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.273,
+      "step": 1622000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2796,
+      "step": 1623000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2758,
+      "step": 1624000
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 5e-05,
+      "loss": 2.2867,
+      "step": 1625000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2816,
+      "step": 1626000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2856,
+      "step": 1627000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2847,
+      "step": 1628000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2873,
+      "step": 1629000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2795,
+      "step": 1630000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2887,
+      "step": 1631000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2748,
+      "step": 1632000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2802,
+      "step": 1633000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2784,
+      "step": 1634000
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 5e-05,
+      "loss": 2.2739,
+      "step": 1635000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2877,
+      "step": 1636000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2779,
+      "step": 1637000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2773,
+      "step": 1638000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.28,
+      "step": 1639000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2827,
+      "step": 1640000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2771,
+      "step": 1641000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2877,
+      "step": 1642000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2777,
+      "step": 1643000
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 5e-05,
+      "loss": 2.2723,
+      "step": 1644000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2762,
+      "step": 1645000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2845,
+      "step": 1646000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2771,
+      "step": 1647000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.278,
+      "step": 1648000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2737,
+      "step": 1649000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2672,
+      "step": 1650000
+    },
+    {
+      "epoch": 1.68,
+      "eval_accuracy": 0.594395614413767,
+      "eval_loss": 2.1171875,
+      "eval_runtime": 4580.2758,
+      "eval_samples_per_second": 138.876,
+      "eval_steps_per_second": 2.17,
+      "step": 1650000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.268,
+      "step": 1651000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2746,
+      "step": 1652000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.2792,
+      "step": 1653000
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 5e-05,
+      "loss": 2.279,
+      "step": 1654000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2716,
+      "step": 1655000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2808,
+      "step": 1656000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2829,
+      "step": 1657000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2804,
+      "step": 1658000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2724,
+      "step": 1659000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2821,
+      "step": 1660000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2768,
+      "step": 1661000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2779,
+      "step": 1662000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2764,
+      "step": 1663000
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 5e-05,
+      "loss": 2.2829,
+      "step": 1664000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2754,
+      "step": 1665000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2814,
+      "step": 1666000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2717,
+      "step": 1667000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2697,
+      "step": 1668000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2825,
+      "step": 1669000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2732,
+      "step": 1670000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2749,
+      "step": 1671000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2757,
+      "step": 1672000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2643,
+      "step": 1673000
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 5e-05,
+      "loss": 2.2784,
+      "step": 1674000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.281,
+      "step": 1675000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2727,
+      "step": 1676000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.285,
+      "step": 1677000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2738,
+      "step": 1678000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2728,
+      "step": 1679000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2696,
+      "step": 1680000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2707,
+      "step": 1681000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2743,
+      "step": 1682000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.264,
+      "step": 1683000
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 5e-05,
+      "loss": 2.2787,
+      "step": 1684000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2674,
+      "step": 1685000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2815,
+      "step": 1686000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2668,
+      "step": 1687000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2783,
+      "step": 1688000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2748,
+      "step": 1689000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2711,
+      "step": 1690000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2724,
+      "step": 1691000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2709,
+      "step": 1692000
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 5e-05,
+      "loss": 2.2737,
+      "step": 1693000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2872,
+      "step": 1694000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2691,
+      "step": 1695000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2724,
+      "step": 1696000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.268,
+      "step": 1697000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2811,
+      "step": 1698000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2744,
+      "step": 1699000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2705,
+      "step": 1700000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2658,
+      "step": 1701000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2686,
+      "step": 1702000
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 5e-05,
+      "loss": 2.2756,
+      "step": 1703000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2761,
+      "step": 1704000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2692,
+      "step": 1705000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2723,
+      "step": 1706000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2698,
+      "step": 1707000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2638,
+      "step": 1708000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2698,
+      "step": 1709000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2736,
+      "step": 1710000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2701,
+      "step": 1711000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2622,
+      "step": 1712000
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 5e-05,
+      "loss": 2.2625,
+      "step": 1713000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2696,
+      "step": 1714000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2676,
+      "step": 1715000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2604,
+      "step": 1716000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2723,
+      "step": 1717000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2704,
+      "step": 1718000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2788,
+      "step": 1719000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2689,
+      "step": 1720000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2688,
+      "step": 1721000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.273,
+      "step": 1722000
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 5e-05,
+      "loss": 2.2755,
+      "step": 1723000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2752,
+      "step": 1724000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2675,
+      "step": 1725000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.27,
+      "step": 1726000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2698,
+      "step": 1727000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2689,
+      "step": 1728000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2633,
+      "step": 1729000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2751,
+      "step": 1730000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2645,
+      "step": 1731000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.269,
+      "step": 1732000
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 5e-05,
+      "loss": 2.2623,
+      "step": 1733000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2674,
+      "step": 1734000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2608,
+      "step": 1735000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2645,
+      "step": 1736000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2743,
+      "step": 1737000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.268,
+      "step": 1738000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2631,
+      "step": 1739000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2689,
+      "step": 1740000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.265,
+      "step": 1741000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2653,
+      "step": 1742000
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 5e-05,
+      "loss": 2.2656,
+      "step": 1743000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2641,
+      "step": 1744000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2687,
+      "step": 1745000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2639,
+      "step": 1746000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2619,
+      "step": 1747000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2727,
+      "step": 1748000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2638,
+      "step": 1749000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2623,
+      "step": 1750000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2589,
+      "step": 1751000
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 5e-05,
+      "loss": 2.2712,
+      "step": 1752000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2608,
+      "step": 1753000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2619,
+      "step": 1754000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2673,
+      "step": 1755000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2657,
+      "step": 1756000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2596,
+      "step": 1757000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2616,
+      "step": 1758000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2713,
+      "step": 1759000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2699,
+      "step": 1760000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2702,
+      "step": 1761000
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 5e-05,
+      "loss": 2.2642,
+      "step": 1762000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2668,
+      "step": 1763000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2629,
+      "step": 1764000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.261,
+      "step": 1765000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2639,
+      "step": 1766000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2629,
+      "step": 1767000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.265,
+      "step": 1768000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2725,
+      "step": 1769000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2676,
+      "step": 1770000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2674,
+      "step": 1771000
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 5e-05,
+      "loss": 2.2598,
+      "step": 1772000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2686,
+      "step": 1773000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2653,
+      "step": 1774000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2672,
+      "step": 1775000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2697,
+      "step": 1776000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2682,
+      "step": 1777000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2587,
+      "step": 1778000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2604,
+      "step": 1779000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2635,
+      "step": 1780000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2573,
+      "step": 1781000
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 5e-05,
+      "loss": 2.2601,
+      "step": 1782000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2717,
+      "step": 1783000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.265,
+      "step": 1784000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2689,
+      "step": 1785000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2669,
+      "step": 1786000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2554,
+      "step": 1787000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2647,
+      "step": 1788000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.259,
+      "step": 1789000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2628,
+      "step": 1790000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2613,
+      "step": 1791000
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 5e-05,
+      "loss": 2.2629,
+      "step": 1792000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.26,
+      "step": 1793000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2547,
+      "step": 1794000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.256,
+      "step": 1795000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2604,
+      "step": 1796000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.269,
+      "step": 1797000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2643,
+      "step": 1798000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2643,
+      "step": 1799000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2658,
+      "step": 1800000
+    },
+    {
+      "epoch": 1.83,
+      "eval_accuracy": 0.5967758944679928,
+      "eval_loss": 2.103515625,
+      "eval_runtime": 4566.2819,
+      "eval_samples_per_second": 139.301,
+      "eval_steps_per_second": 2.177,
+      "step": 1800000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2635,
+      "step": 1801000
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 5e-05,
+      "loss": 2.2641,
+      "step": 1802000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2638,
+      "step": 1803000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2513,
+      "step": 1804000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.264,
+      "step": 1805000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2626,
+      "step": 1806000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2563,
+      "step": 1807000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2593,
+      "step": 1808000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2606,
+      "step": 1809000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2542,
+      "step": 1810000
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 5e-05,
+      "loss": 2.2531,
+      "step": 1811000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2613,
+      "step": 1812000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2553,
+      "step": 1813000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2647,
+      "step": 1814000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2572,
+      "step": 1815000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2569,
+      "step": 1816000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2603,
+      "step": 1817000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2624,
+      "step": 1818000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2526,
+      "step": 1819000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2613,
+      "step": 1820000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 5e-05,
+      "loss": 2.2551,
+      "step": 1821000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2556,
+      "step": 1822000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2519,
+      "step": 1823000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2496,
+      "step": 1824000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2647,
+      "step": 1825000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2559,
+      "step": 1826000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2538,
+      "step": 1827000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2623,
+      "step": 1828000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2558,
+      "step": 1829000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2518,
+      "step": 1830000
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 5e-05,
+      "loss": 2.2601,
+      "step": 1831000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2527,
+      "step": 1832000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2587,
+      "step": 1833000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2674,
+      "step": 1834000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2587,
+      "step": 1835000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2532,
+      "step": 1836000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2586,
+      "step": 1837000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2579,
+      "step": 1838000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2581,
+      "step": 1839000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2645,
+      "step": 1840000
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 5e-05,
+      "loss": 2.2569,
+      "step": 1841000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2639,
+      "step": 1842000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2627,
+      "step": 1843000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2643,
+      "step": 1844000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2574,
+      "step": 1845000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.256,
+      "step": 1846000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2574,
+      "step": 1847000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2504,
+      "step": 1848000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2543,
+      "step": 1849000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2613,
+      "step": 1850000
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 5e-05,
+      "loss": 2.2614,
+      "step": 1851000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2564,
+      "step": 1852000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2655,
+      "step": 1853000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2476,
+      "step": 1854000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2561,
+      "step": 1855000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2623,
+      "step": 1856000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2472,
+      "step": 1857000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.261,
+      "step": 1858000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2546,
+      "step": 1859000
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 5e-05,
+      "loss": 2.2587,
+      "step": 1860000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2576,
+      "step": 1861000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2536,
+      "step": 1862000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2527,
+      "step": 1863000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2592,
+      "step": 1864000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.258,
+      "step": 1865000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2542,
+      "step": 1866000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2517,
+      "step": 1867000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2614,
+      "step": 1868000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2525,
+      "step": 1869000
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 5e-05,
+      "loss": 2.2488,
+      "step": 1870000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2509,
+      "step": 1871000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2642,
+      "step": 1872000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2598,
+      "step": 1873000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2629,
+      "step": 1874000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2537,
+      "step": 1875000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2517,
+      "step": 1876000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2579,
+      "step": 1877000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2457,
+      "step": 1878000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2506,
+      "step": 1879000
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 5e-05,
+      "loss": 2.2564,
+      "step": 1880000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2564,
+      "step": 1881000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2564,
+      "step": 1882000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2536,
+      "step": 1883000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2429,
+      "step": 1884000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2544,
+      "step": 1885000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2491,
+      "step": 1886000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2527,
+      "step": 1887000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2486,
+      "step": 1888000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2482,
+      "step": 1889000
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 5e-05,
+      "loss": 2.2472,
+      "step": 1890000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2465,
+      "step": 1891000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2472,
+      "step": 1892000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2524,
+      "step": 1893000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2592,
+      "step": 1894000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2507,
+      "step": 1895000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2486,
+      "step": 1896000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2528,
+      "step": 1897000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2456,
+      "step": 1898000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2582,
+      "step": 1899000
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5e-05,
+      "loss": 2.2483,
+      "step": 1900000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2546,
+      "step": 1901000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2491,
+      "step": 1902000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2475,
+      "step": 1903000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2487,
+      "step": 1904000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2428,
+      "step": 1905000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2524,
+      "step": 1906000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2587,
+      "step": 1907000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.253,
+      "step": 1908000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.255,
+      "step": 1909000
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5e-05,
+      "loss": 2.2485,
+      "step": 1910000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2547,
+      "step": 1911000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.252,
+      "step": 1912000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2541,
+      "step": 1913000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2504,
+      "step": 1914000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2501,
+      "step": 1915000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2462,
+      "step": 1916000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2562,
+      "step": 1917000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2438,
+      "step": 1918000
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5e-05,
+      "loss": 2.2566,
+      "step": 1919000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2463,
+      "step": 1920000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.253,
+      "step": 1921000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2491,
+      "step": 1922000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2505,
+      "step": 1923000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2595,
+      "step": 1924000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.242,
+      "step": 1925000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2472,
+      "step": 1926000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2541,
+      "step": 1927000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2577,
+      "step": 1928000
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5e-05,
+      "loss": 2.2409,
+      "step": 1929000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2457,
+      "step": 1930000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2485,
+      "step": 1931000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2388,
+      "step": 1932000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.238,
+      "step": 1933000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2509,
+      "step": 1934000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.242,
+      "step": 1935000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.253,
+      "step": 1936000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2549,
+      "step": 1937000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2496,
+      "step": 1938000
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5e-05,
+      "loss": 2.2539,
+      "step": 1939000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2536,
+      "step": 1940000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2491,
+      "step": 1941000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2433,
+      "step": 1942000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2435,
+      "step": 1943000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2485,
+      "step": 1944000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2504,
+      "step": 1945000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2404,
+      "step": 1946000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2496,
+      "step": 1947000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.2498,
+      "step": 1948000
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5e-05,
+      "loss": 2.245,
+      "step": 1949000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2511,
+      "step": 1950000
+    },
+    {
+      "epoch": 1.99,
+      "eval_accuracy": 0.5974179462628302,
+      "eval_loss": 2.09375,
+      "eval_runtime": 4575.295,
+      "eval_samples_per_second": 139.027,
+      "eval_steps_per_second": 2.172,
+      "step": 1950000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2454,
+      "step": 1951000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2447,
+      "step": 1952000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.246,
+      "step": 1953000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2494,
+      "step": 1954000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2503,
+      "step": 1955000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2404,
+      "step": 1956000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2426,
+      "step": 1957000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2461,
+      "step": 1958000
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5e-05,
+      "loss": 2.2424,
+      "step": 1959000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 2.2524,
+      "step": 1960000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 2.2414,
+      "step": 1961000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 2.2425,
+      "step": 1962000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 2.2458,
+      "step": 1963000
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5e-05,
+      "loss": 2.2477,
+      "step": 1964000
+    },
+    {
+      "epoch": 2.0,
+      "step": 1964054,
+      "total_flos": 2.7512471571964887e+19,
+      "train_loss": 0.10199262492233857,
+      "train_runtime": 71458.7973,
+      "train_samples_per_second": 3518.093,
+      "train_steps_per_second": 27.485
+    }
+  ],
+  "max_steps": 1964054,
+  "num_train_epochs": 2,
+  "total_flos": 2.7512471571964887e+19,
+  "trial_name": null,
+  "trial_params": null
+}